Activity 4

library(tidyverse)

Registered S3 methods overwritten by 'dbplyr':
  method         from
  print.tbl_lazy     
  print.tbl_sql      
── Attaching packages ──────────────────────────────────────────────────────────────────────────────── tidyverse 1.3.1 ──
✓ ggplot2 3.3.3     ✓ purrr   0.3.4
✓ tibble  3.1.2     ✓ dplyr   1.0.6
✓ tidyr   1.1.3     ✓ stringr 1.4.0
✓ readr   1.4.0     ✓ forcats 0.5.1
── Conflicts ─────────────────────────────────────────────────────────────────────────────────── tidyverse_conflicts() ──
x dplyr::filter() masks stats::filter()
x dplyr::lag()    masks stats::lag()

library(modelr)

set.seed(64)

Data Preperation

heights$income <- heights$income/1000
heights <- heights[-c(721,973,1611,1641,1866,2828,3122,6178,6319,6794),]
heights

Split Data into 6 Cross Validation Intervals

cv  <- crossv_kfold(heights, k=6)
cv

Fit Models to each training set

models1  <- map(cv$train, ~lm(income ~ height, data = .))
models2  <- map(cv$train, ~lm(income ~ sex, data = .))
models3  <- map(cv$train, ~lm(income ~ education, data = .))

Get Predictions using the models

get_pred <- function(model, test_data) {
  data <- as.data.frame(test_data)
  pred <- add_predictions(data, model)
  return(pred)
}

pred1  <- map2_df(models1, cv$test, get_pred, .id = "Run")
pred2  <- map2_df(models2, cv$test, get_pred, .id = "Run")
pred3  <- map2_df(models3, cv$test, get_pred, .id = "Run")

calculate the MSE for each group

MSE1  <- pred1 %>% group_by(Run) %>% 
  summarise(MSE = mean( (income - pred)^2))
MSE1

MSE2  <- pred2 %>% group_by(Run) %>% 
  summarise(MSE = mean( (income - pred)^2))
MSE2

MSE3  <- pred3 %>% group_by(Run) %>% 
  summarise(MSE = mean( (income - pred)^2))
MSE3

Compare the models

paste("Model 1 MSE:", mean(MSE1$MSE))

[1] "Model 1 MSE: 2978.99924246456"

paste("Model 2 MSE:", mean(MSE2$MSE))

[1] "Model 2 MSE: 2984.04148812215"

paste("Model 3 MSE:", mean(MSE3$MSE))

[1] "Model 3 MSE: 2645.51716314473"